Accurate breast cancer screening demands that radiologists maintain a balance of high sensitivity and high specificity when interpreting mammography. Subspecialty-trained breast imagers perform significantly better than general radiologists by recognizing more breast cancers and minimizing benign biopsies. An automated reasoning system called a Bayesian network (BN), has the potential to improve the performance of general radiologists, who interpret the majority of mammograms, to the level of subspecialty-trained breast radiologists, who are in short supply. A BN is a probabilistic graphical model that has been used for decision support in a variety of domains, including radiology. Machine learning provides an appealing way to create and optimize BNs to perform at high levels of sensitivity and specificity. Our research group has developed a prototype, expert-defined BN that uses imaging features and demographic risk factors to classify abnormalities on mammograms as benign or malignant. Our BN can perform at a higher level than general radiologists, but our goal is for it to perform as well or better than subspecialty-trained breast radiologists. To this end, we have compiled a structured, multi-relational dataset of mammography abnormalities and pathologic outcomes from which cutting edge machine learning techniques can construct an improved BN. In the past, BNs were trained and tested on individual abnormalities in isolation. In fact, other abnormalities on the same mammogram or previous mammograms, which necessarily appear in other rows of a relational database, can further improve BN learning. We have used basic statistical relational learning (SRL) techniques to enhance Bayesian learning algorithms to leverage this important additional data. A novel SRL capability introduced by our team within the last year, called view learning, makes it possible to incorporate this related data from other parts of the database by automatically defining new database fields. In our preliminary work, we have shown a stepwise improvement in BN performance: first, with conventional BN learning;then, with basic SRL;and finally, with view learning. We now seek support to determine whether more tightly integrated SRL and view learning will significantly improve our BN's ability to accurately diagnose breast cancer. In addition, we propose to investigate whether two other promising machine learning techniques, predicate-invention and collective-classification, can optimize the BN to perform at levels significantly better than current clinical practice.

Agency
National Institute of Health (NIH)
Institute
National Cancer Institute (NCI)
Type
Research Project (R01)
Project #
3R01CA127379-03S1
Application #
7919663
Study Section
Biodata Management and Analysis Study Section (BDMA)
Program Officer
Croft, Barbara
Project Start
2009-08-01
Project End
2011-12-31
Budget Start
2009-08-01
Budget End
2011-12-31
Support Year
3
Fiscal Year
2009
Total Cost
$296,245
Indirect Cost
Name
University of Wisconsin Madison
Department
Radiation-Diagnostic/Oncology
Type
Schools of Medicine
DUNS #
161202122
City
Madison
State
WI
Country
United States
Zip Code
53715
Feld, Shara I; Fan, Jun; Yuan, Ming et al. (2018) Utility of Genetic Testing in Addition to Mammography for Determining Risk of Breast Cancer Depends on Patient Age. AMIA Jt Summits Transl Sci Proc 2017:81-90
Li, Hui; Zhu, Yitan; Burnside, Elizabeth S et al. (2016) MR Imaging Radiomics Signatures for Predicting the Risk of Breast Cancer Recurrence as Given by Research Versions of MammaPrint, Oncotype DX, and PAM50 Gene Assays. Radiology 281:382-391
Abbey, Craig K; Wu, Yirong; Burnside, Elizabeth S et al. (2016) A Utility/Cost Analysis of Breast Cancer Risk Prediction Algorithms. Proc SPIE Int Soc Opt Eng 9787:
Burnside, Elizabeth S; Liu, Jie; Wu, Yirong et al. (2016) Comparing Mammography Abnormality Features to Genetic Variants in the Prediction of Breast Cancer in Women Recommended for Breast Biopsy. Acad Radiol 23:62-9
Benndorf, Matthias; Wu, Yirong; Burnside, Elizabeth S (2016) A history of breast cancer and older age allow risk stratification of mammographic BI-RADS 3 ratings in the diagnostic setting. Clin Imaging 40:200-4
Fan, Jun; Wu, Yirong; Yuan, Ming et al. (2016) Structure-Leveraged Methods in Breast Cancer Risk Prediction. J Mach Learn Res 17:
Li, Hui; Zhu, Yitan; Burnside, Elizabeth S et al. (2016) Quantitative MRI radiomics in the prediction of molecular classifications of breast cancer subtypes in the TCGA/TCIA data set. NPJ Breast Cancer 2:
Wu, Yirong; Abbey, Craig K; Liu, Jie et al. (2016) Discriminatory power of common genetic variants in personalized breast cancer diagnosis. Proc SPIE Int Soc Opt Eng 9787:
Wu, Yirong; Liu, Jie; Del Rio, Alejandro Munoz et al. (2015) Developing a clinical utility framework to evaluate prediction models in radiogenomics. Proc SPIE Int Soc Opt Eng 9416:
Liu, Jie; Wu, Yirong; Ong, Irene et al. (2015) Leveraging Interaction between Genetic Variants and Mammographic Findings for Personalized Breast Cancer Diagnosis. AMIA Jt Summits Transl Sci Proc 2015:107-11

Showing the most recent 10 out of 43 publications